The purpose of the project is to assess the evidence for an association between parental alcohol use and children's injuries in two representative samples of U.S. children using data from the 1985 and 1988 National Health Interview Surveys. We also aim to assess social, psychological and life style characteristics which may constitute mechanisms linking alcohol use to children's injuries. Alcohol use is associated with motor vehicle injuries, burns, and drownings in the person sustaining the injury. While it seems likely that children in the care of intoxicated adults are at higher risk of injury than others, evidence for the relationship between parental use of alcohol and children's injuries is seriously lacking. The 1985 National Health Interview Survey (NHIS) included a supplement on Health Promotion and Disease Prevention which contained items on alcohol behavior, use and knowledge of injury preventing strategies. The 1988 NHIS included a supplement with extensive information on adult alcohol use and a child health supplement with detailed data on children's injuries. In the first year of the project, assessments will be made of the association between parental alcohol use from the 1985 Supplement and injuries reported in the two week period preceding the interview. In the second year, replication of the findings from the first study will be attempted using the 1988 data, and will be extended to focus of severity and circumstances of injuries reported in the two-week period and in the year preceding the interview. We hypothesize that: 1) there will be a dose-response relationship between children's injuries and the intensity of parental alcohol use; 2) indicators of impaired family functioning will mediate the relationship between alcohol injuries, but that parental use of alcohol will also have a direct effect on children's injuries; 3) health promoting behaviors and use of injury control strategies will mediate the association between alcohol and children's injuries, but that parental use of alcohol will also have a direct effect on children's injuries; 4) problem behavior in the children of alcohol users will mediate the association, but that parental use of alcohol will also have a direct effect on children's injuries. Multiple regression and logistic regression will be used to test these hypotheses. Evidence from a national study can be used to support social and legislative efforts to control the impact of alcohol on the health of children.